This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: __eprintf again
- To: Ted Irons <ironst at aries dot tucson dot saic dot com>
- Subject: Re: __eprintf again
- From: Jeffrey A Law <law at cygnus dot com>
- Date: Tue, 23 Jun 1998 23:41:03 -0600
- cc: egcs at cygnus dot com
- Reply-To: law at cygnus dot com
In message <359002D1.702A@aries.tucson.saic.com>you write:
> Sorry - my previous posting was incomplete.
>
> On a Solaris 2.5.1 box,
> while trying to build the 19980619b snapshot of the blitz library
> ( http://monet.uwaterloo.ca/blitz/ )
> with egcs-19980621 and binutils-2.9.1 I get the following
> error:
>
> g++ -I./.. -ftemplate-depth-30 -g -DBZ_DEBUG -c ./ctors.cpp
> g++ ctors.o -o ctors -L../lib -lblitz -lm
> Undefined first referenced
> symbol in file
> __eprintf ctors.o
> ld: fatal: Symbol referencing errors. No output written to ctors
> collect2: ld returned 1 exit status
> gmake[1]: *** [ctors] Error 1
> gmake: *** [check] Error 2
>
>
> Doing nm on libgcc.a comes up with
> _eprintf.o:
Hmmm, did it list any symbols for eprintf.o?
Did you configure with the "--with-newlib" option? That disables
__eprintf in libgcc.a. If yes, then that's where you went wrong :-)
since newlib is for embedded systems, not native ports.
jeff